Summary: An MRI feedback method induced specific feelings about people’s faces, new research reports.
Source: Brown University.
Volunteers who began the experiment feeling neutral about certain faces later adopted feelings that researchers induced using an MRI feedback technique, without their awareness, according to a newly published study.
The research, published in PLOS Biology, indicates that a single brain region can host both positive and negative reactions to faces. It also provides a second demonstration that an MRI-based training method can shape a mental process in participants who do not know what is being trained. Earlier this year, the same team used the technique to couple the perception of a color with a visual pattern so effectively that volunteers reported seeing the color when the pattern appeared, even if the color was absent.
In the current study, the researchers tested whether they could influence feelings about faces — a more complex social function — with the long-term goal of developing the approach into a therapeutic tool, for example to help treat anxiety.
“Face recognition is a very important social function for people,” said co-author Takeo Watanabe, the Fred M. Seed Professor of Cognitive and Linguistic Sciences at Brown University. “Facial recognition is associated with people’s emotions.”
Decoded neurofeedback explained
The method the team used, called DecNef (decoded neurofeedback), begins by identifying the specific patterns of brain activity that correspond to a particular mental state. In this study, 24 volunteers first viewed hundreds of faces and rated their feelings about each one on a scale from 1 (dislike) to 10 (like), with 5 representing neutral. While participants made those ratings, researchers recorded activity in a brain area called the cingulate cortex using fMRI.
Many neuroscientists have assumed that positive and negative feelings about faces arise in different parts of the brain. The Brown–Kyoto team, however, looked for distinct activation patterns within the cingulate cortex itself. Their analysis software, a decoder, was able to identify consistent patterns in each volunteer’s cingulate cortex associated with liking or disliking faces.
“We found that the cingulate cortex seems to handle both opposing directions with different activity patterns,” said co-author Yuka Sasaki, associate professor (research) of cognitive, linguistic and psychological sciences at Brown.
After mapping these signature patterns for each participant, the researchers divided the volunteers into two groups of 12 — a positive group and a negative group — without telling them which group they were in. Over a series of MRI sessions on subsequent days, participants viewed a subset of faces they had rated as neutral. Immediately after each face, they saw a disk on the screen and were asked to try to make it appear as large as possible, with a larger disk tied to a slightly higher monetary reward.
Unknown to the volunteers, the only way the disk would grow was if the MRI patterns recorded at that moment matched the participant’s pre-identified signature pattern for either positive or negative feeling. In other words, members of the positive group were rewarded with a larger disk when their brain activity matched the pattern associated with liking a face after seeing a neutral face; members of the negative group were rewarded when their activity matched the pattern associated with disliking the face.
DecNef therefore trains participants to produce particular internal states by rewarding the moments when those states occur, even though the participant is not consciously aware of what is being rewarded. A separate control group of six volunteers rated faces but did not take part in the DecNef training involving the disk task.
After the training sessions, all participants were asked again to rate the faces they had previously judged as neutral.
Facial feelings were affected
Analysis of the results produced several clear findings. On average, ratings from the group trained toward positive feelings rose modestly but significantly — by about 0.6 points on the 1–10 scale. Ratings from the negative group declined slightly but significantly. The control group’s ratings showed no meaningful change.
The authors conclude that associating originally neutral faces with covert induction of specific activation patterns in the cingulate cortex led to directed changes in facial preference: the same brain region, when driven into distinct patterns, produced opposite emotional responses toward the same faces.
To confirm that the change was not deliberate, the researchers interviewed participants after the experiment. None reported understanding the true purpose of the disk task. When asked to guess whether they had been placed in the positive or negative group, participants performed no better than chance. This supports the conclusion that changes in preference occurred without conscious intention.
Further analysis showed a strong correlation (0.78) between the degree to which the cingulate cortex exhibited the target activation pattern during training and the magnitude of the subsequent change in preference. In other words, a greater expression of the targeted pattern predicted a larger shift in how a face was rated.
Toward a DecNef therapy
The induced shifts in feeling were modest, but the training took place over only a few days. The researchers note that longer training, on the scale of weeks as is common in clinical therapy, might produce larger effects.
Even a small reduction in an aversive response can be meaningful. “If someone develops a traumatic memory that makes him or her suffer, even a small reduction of the suffering would be helpful,” Watanabe said.
The team also acknowledged ethical concerns. Because the method can change preferences without awareness, there is potential for misuse, so limitations and safeguards will be important as the approach develops.
The paper’s authors are Kazuhisa Shibata (lead author), Takeo Watanabe, Mitsuo Kawato (corresponding author), and Yuka Sasaki.
Source: David Orenstein, Brown University
Image Source: NeuroscienceNews.com image adapted from the Brown University press release.
Original Research: Differential Activation Patterns in the Same Brain Region Led to Opposite Emotional States, PLOS Biology. Published online September 8, 2016. doi:10.1371/journal.pbio.1002546
Abstract
Differential Activation Patterns in the Same Brain Region Led to Opposite Emotional States
This study used fMRI decoded neurofeedback (DecNef) to test whether different multi-voxel activation patterns within the cingulate cortex represent opposite directions of facial preference. While neutrally preferred faces were presented, DecNef repeatedly induced activation patterns corresponding to higher or lower preference. As a result, faces that were initially neutral became more or less preferred, depending on the pattern induced. The findings indicate that distinct activation patterns within the same brain region, rather than averaged activity in different areas, can determine positive or negative facial preference. This approach highlights the importance of activation patterns within brain regions for cognitive functions.